The Influence of Sonication on the Thermal Behavior of Muscovite And

Journal of the European Ceramic Society 24 (2004) 2793–2801 www.elsevier.com/locate/jeurceramsoc

The inﬂuence of sonication on the thermal behavior of muscovite and biotite

Luis A. Pe´ rez-Maquedaa,*, Jose´ M. Blanesa, Jose´ Pascualb, Jose´ L. Pe´ rez-Rodrı´ gueza aInstituto de Ciencia de Materiales de Sevilla, CSIC, Universidad de Sevilla, Ame´rico Vespucio s/n 41092 Sevilla, Spain bDepartamento de Ingenierı´a Civil, Materiales y Fabricacio´n, ETSII, Universidad de Ma´laga, Plaza, Ejido s/n 29013 Ma´laga, Spain

Received 2 November 2002; accepted 9 April 2003

Abstract The differences on the thermal behavior (DTA-TG) of mica samples measured before and after sonication have been studied. Sonication treatment produces important modifications in the thermal behavior of muscovite and biotite samples. For muscovite, it produces a broadening and decrease in temperature of the dehydroxylation and crystallization effects, reaching a steady stage after 40 h treatment time. For biotite, the original single peak profile for the dehydroxylation of the untreated sample is converted into a two peaks profile after sonication, the intensity of the low temperature peak increases with sonication time, while the intensity of the high temperature peak decreases. The modification of the thermal behavior for sonicated samples has been correlated to the particle size distribution effect produced by the sonication treatment. It has been also observed that the cup tip of the sonication equipment contaminates the samples releasing titanium of its composition. # 2003 Elsevier Ltd. All rights reserved. Keywords: Micas; Particle size; Sonication; Thermal properties

1. Introduction been explained by considering that the Al–OH bond is greatly affected by the coordination number of neigh- Micas are minerals of significant commercial impor- boring polyhedron. When neighboring polyhedron are tance due to their physical properties. These minerals in octahedral coordination, the hydroxyl group is lost remain stable at high temperature and have interesting at lower temperatures than when neighboring poly- electrical properties. Therefore, they are used in appli- hedron are in five-fold coordination (after partial cations where it is required high-temperature stability dehydroxylation). and electrical properties.1 Mechanical treatment are of great importance in the The thermal behavior of micas has been extensively preparation and processing of raw materials.5 Micas are studied. Thus, studies on the dehydroxylation of dioc- often used in ground forms. Mica may be dry ground tahedral micas, such as celadonite and glauconite, have (yielding particles in the range from 1.2 mm to 150 mm), shown that during dehydroxylation cations migrate wet ground (95–45 mm), or micronized (<53 mm). When from cis into trans-octahedra and have five-fold coordi- studying grinding treatments emphasis has been placed nation.2 For muscovite, Mackenzie et al.3 have sug- on the impact of milling on the physicochemical prop- gested an homogeneous dehydroxylation mechanism erties of the initially coarse powder, and its influence on forming a dehydroxylate in which the aluminum is pre- the transformation of the solid.6,7 It is well known that dominantly 5-coordinate. On the other hand Guggen- grinding of clay minerals produces various effects on their heim et al.4 have found a very broad peak in the structure and properties.8 The significant processes differential thermogravimetric (DTG) trace of a mica involved in the preparation of ceramic raw materials have sample that they interpreted as two overlapping poorly been extensively studied, specially those for kaolinite,9,10 resolved dehydroxylation peaks. These two peaks have montmorillonite or bentonite,11 illite,12 pyrophyllite,13 talc,14 and vermiculite.15 * Corresponding author. It has been observed that grinding of clays produces E-mail address: [email protected] (L.A. Pe´ rez-Maqueda). progressive amorphyzation when grinding time increa-

0955-2219/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.jeurceramsoc.2003.10.002 2794 L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 ses.16 In many cases this treatment yields heterogeneous 2.2. Sonication hardly aggregated materials with modified chemical reactivity. All these effects are useful for some processes, A high-intensity ultrasonic horn (Misonix inc.), that but they are important drawbacks for many of the consists of a solid titanium rod connected to a piezo- applications of mica, where it is important to have small electric ceramic, and a 20 KHz, 750 W power supply particle diameters while maintaining the crystalline were used. The horn tip was introduced into a structure in order to retain the properties of mica. An thermostated (20 C) double-jacket reactor. Three alternative method for particle size reduction is sonica- grams of <2 mm flakes were mixed with 100 cm3 of tion. Cavitational collapse of bubbles on solid surfaces freshly deionized water and subjected to ultrasounds for leads to micro jet and shock-wave impacts on the sur- periods ranging between 10 and 100 h. face of the solids, together with interparticle collisions which can result in particle size reduction.17,18 Recently, 2.3. X-ray diffraction analysis micron and submicron-sized vermiculite flakes were prepared from a natural macroscopic vermiculite sam- Diffraction patterns were obtained using a dif- ple by sonication.17,19 The resulting materials were fractometer (Kristalloflex D-500 Siemens) at 36 kV and crystalline, as assayed by X-ray diffraction.20 26 mA with Ni-filtered CuKa radiation and a graphite Particle size distribution is a very sensitive issue in the monochromator. thermal behavior of materials. The importance of this issue has been recognized in the literature and authors 2.4. Nuclear magnetic resonance measurements have observed changes in the profiles of the thermal transformation of solid when performing experiments High-resolution solid-state 29Si and 27Al magic angle under both isothermal and non-isothermal con- spinning (MAS) nuclear magnetic resonance (NMR) ditions.21À24 For kinetically driven process, it has been spectra of powdered samples were recorded at 79.49 and shown in a theoretical work22 that the shape and aver- 104.26 MHz, respectively, in a Bru¨ ker MSL-400 spec- age temperature of the thermogravimetric curves are trometer. Measurements were conducted at room tem- affected by the particle size distribution. This change does perature with tetramethylsilane (TMS) and 3+ not necessarily imply a modification in the kinetic para- [Al(H2O)6] as external references. meters (activation energy, pre-exponential parameter of Arrhenius, kinetic model) of the reaction. 2.5. Thermal study Clay minerals subjected to grinding experience particle size reduction together with other important trans- Thermogravimetric analysis (TG), differential ther- formations, such as amorphyzation, aggregation of mogravimetric analysis (DTG) and differential thermal small particles into larger units and contamination. The analysis (DTA) were carried out simultaneously in static influence of grinding on the thermal behavior of miner- air or argon flow (500 cm3 minÀ1) with an automatic als has been extensively studied.16,25À28 However, until thermal analyzer system (Seiko, TG/DTA 6300). Mica now, no consideration has been given to the effect of (muscovite and biotite) samples of about 50 mg were sonication on the thermal behavior of minerals. The loosely packed into a platinum holder and were ther- purpose of this work is to show the effect of sonication mally treated at a heating rate of 10 C minÀ1. on the thermal behavior of muscovite and biotite, spe- cial attention is given to the effect of particle size dis- 2.6. Scanningelectron microscope study tribution. In addition, the influence of the possible contamination induced by the tip cups of the ultrasound The samples were studied by scanning electron equipment on the thermal analysis and new phases pro- microscopy (SEM) using a Jeol JSM-5400 model. The duced is investigated. samples were covered with a thin gold film and analyzed by energy dispersive X-ray spectroscopy (EDX) using a Link-ISIS Si/Li detector. The mean size (length, L) of 2. Experimental the plate-like particles was evaluated by measuring the length in the longest direction of around 100 particles 2.1. Materials for each sample.

A muscovite from Fuente Obejuna (Co´ rdoba, Spain) 2.7. Particle size distribution and a biotite from Santa Olalla (Huelva, Spain) were used as starting materials.29 Both samples consist of pla- The mass fraction versus particle diameter plots have telets of about 10 cm in length and 0.5 cm in thickness. been obtained by a deposition–centrifugation separa- Previously to the sonication treatment, samples were tion procedure described in the literature30 followed by lightly ground using a knife-mill and sieved under 2 mm. the gravimetric quantiﬁcation of the diﬀerent fractions. L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 2795

2.8. Chemical analysis intensity of the diffractions due to textural effects.33 These results indicate that the structure of the micas is Chemical quantification of Ti in bulk samples was maintained after the sonication treatment in contrast carried out using a sequential X-ray fluorescence with those results previously obtained for ground clay spectrometer (Siemens, SRS 3000). minerals,9,13,16,25À27 where even short milling treatments produce significant structure modifications. The gradual size reduction on sonication of muscovite 3. Results and biotite was revealed by SEM (Fig. 1). During sonication treatment, the original stacking layers of musco- The 27Al and 29Si NMR analysis of the untreated vite and biotite are delaminated and the lamellar and sonicated muscovite samples (figures not shown) phyllosilicate mineral particles are broken by mechan- did not show at the atomic scale any modification in the ical impact induced by sonication producing a decrease coordination of aluminum or silicon.29 Thus, for both in particle size. The evolution of the particle length with sonicated and untreated samples, the 27Al spectra sonication time (Fig. 2) shows for the muscovite an showed the typical bands corresponding to aluminum in important decrease of particle length with sonication octahedral and tetrahedral sites3,31 and the 29Si spectra time up to a limit reached for 40 h. For the biotite, the 3,32 showed a peak attributed to Si(Si2Al) environment. particle length decreases in the entire sonication range Additionally, the X-ray diffraction patterns of the studied here. The particle lengths of the sonicated original and sonicated muscovite and biotite samples samples are larger for biotite than for muscovite. indicated that sonication does not produce significant The mass fraction as a function of the particle dia- structural changes in the samples. Thus, the diffraction meter plots for the muscovite and biotite samples soni- lines of both untreated and sonicated samples remained cated for 40 h are presented in Fig. 3. These plots show unchanged, except for the broadening of the lines, that that the sonicated samples are not monodispersed but could be attributed to the delamination and crystallite constituted of different particle size fractions in the size reduction,17 and to the change in the relative range of micron and submicron sizes. It is also clear from this figure, as it was also revealed from SEM measurements, that the sonicated biotite sample has a larger particle size than the muscovite one. DTG diagrams carried out in argon flow of untreated and sonicated muscovite samples in the range from 300 to 1000 C, where dehydroxylation of muscovite takes place, are shown in Fig. 4. The weight loss of the

Fig. 1. Scanning electron micrographs of muscovite (a) and biotite (b) Fig. 2. Evolution of the mean size (length) as a function of the samples sonicated for 40 h. sonication time for biotite (&) and muscovite (*). 2796 L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 untreated sample begins at 650 C and reaction is over size distribution. The importance of particle size in the at 920 C. The total weight loss agrees with the theore- thermal reactivity of solids have been recognized in lit- tical one calculated for an ideal muscovite (4.70%). erature.21À24 Theoretical studies22 have shown that, for This theoretical weight loss is assumed to be due entirely kinetically controlled processes, the shape of the TG- to the structural water lost by dehydroxylation as esti- DTG traces and the average temperature of the process mated from the crystal structure. Sonication modifies are affected by the particle size distribution. Thus, in the DTG profile (Fig. 4). The relatively sharp weight two identical samples with different particle size dis- loss centered at 824 C for the untreated sample is contribution, the differences in their thermal behavior could verted into a much broader weight loss shifted to lower be attributed to the particle size effect and not necessa- temperatures. Thus, the DTG maximum for the sample rily to a change in the kinetic mechanism or in the sonicated for 10 h (Fig. 4b) appears at 700 C (125 C kinetic parameters (activation energy or pre-exponential lower than for the untreated sample). Longer sonication factor of Arrhenius). times produce even a more pronounced shifting of this For the muscovite sample, there is a correlation DTG peak. This shifting effect reaches a limit for 40 h between the effect of sonication on particle size reduc- sonication time (maximum at 640 C, Fig. 4c). In addition and the changes in the thermal behavior. Thus, the tion to this weight loss, a new weight loss is observed for thermal behavior changes while the particle size is the sonicated samples in the range from 300 to 450 C reduced, that is up to 40 h treatment time, remaining that appears as a shoulder in the DTG trace (Fig. 4b–d). unchanged for longer treatments times. The particle size This weight loss also increases with sonication time up reduction produced by sonication facilitates the dehy- to a limit reached for 40 h treatment time. droxylation of the muscovite as observed by the shifting Since X-ray diffraction and NMR studies have shown of the main weight loss toward lower temperature. In that sonication does not produce significant changes in addition, the new borders created by the particle size the mica crystalline structure or Al and Si coordination, reduction should be responsible of the low temperature the effect of sonication on the thermal behavior of the (300–450 C) due to weakly bonded OH groups placed muscovite could be attributed to changes in the particle on these borders.

Fig. 4. DTG curves obtained under argon ﬂow at a heating rate of Fig. 3. Mass fraction as a function of the particle diameter for the 10 CmÀ1 for the muscovite before (a) and after sonication for 10 h muscovite (a) and biotite (b) samples sonicated for 40 h. (b), 40 h (c), and 100 h (d). L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 2797

In order to study the influence of the particle size dis- mullite, as was confirmed by X-ray diffraction of the tribution on the thermal behavior of the sonicated sonicated muscovite samples heated to 1250 C (figure muscovite, DTG plots of the different fractions with not shown). The products of recrystallization may vary different particle sizes separated by the deposition–cen- somewhat for different natural muscovite samples.34 In trifugation procedure were recorded for the sample Fig. 6b, it has been included the DTA of the sample sonicated for 40 h under argon atmosphere. Fig. 5 sonicated for 100 h. For this sample, the dehydroxylation shows the DTG traces for the different size fractions effect is also broadened and shifted to lower tempera- after normalization and multiplication by the amount of tures as compared with that of the untreated material. sample as taken from Fig. 3, the overall curve resulted The evolution of this peak is identical to that observed of adding up of the DTG traces of the different fractions for the DTG trace (Fig. 4). The DTA effect at tempera- is also included (solid line). It is clear from this figure ture higher than 900 C suffers also broadening and (Fig. 5) that the overall trace resulting of adding up the shifting to lower temperature with the sonication time. contribution of all the fractions has the same shape as The shifting of the higher temperature endothermic the DTG trace shown in Fig. 4b for the muscovite peak has also been observed for ground muscovite.34 sonicated for 40 h. Fig. 5 allows to discriminate the Fig. 6c shows the DTA diagram recorded in static air influence of the different particle sizes on the DTG for the sonicated muscovite sample. The endothermic trace. Thus, the fractions with smaller particle size have effects described above for the experiments performed in a more siginificant contribution to the low temperature argon are also observed in the DTA traces recorded in weight loss than the larger particles. In addition, the air. However, an additional exothermic effect at 682 C main weight loss with a maximum in the DTG at 640 C is present in treated sample. TG experiments (figure not is the result of the contributions of all the fractions in shown) correlate this exothermic effect with an increase such a way that the smaller particles dehydroxylate at in weight at the same temperatures. This exothermic lower temperature than the larger ones. The broadening effects and the corresponding weight increase may be of the DTG peak of the sonicated samples is also due to only attributed to oxidation and/or nitridation of some the contribution of the different particle size fractions. These results illustrate the effect that the changes in the mean particle size and particle size distribution have on the thermal dehydroxylation of muscovite. The DTA curve of the original muscovite carried out in argon flow (Fig. 6a) shows two endothermic peaks, one with a minimum at 825 C and the other with a minimum at 1145 C. The temperature of former endothermic peak matches that of the weight loss in the DTG curve (Fig. 4a) and corresponds to dehydroxylation. The endothermic peak at higher temperatures represent recrystallization into leucite, corundum and

Fig. 5. DTG traces of the different fractions with different particle sizes separated by the deposition–centrifugation procedure recorded for the muscovite sample sonicated for 40 h under argon flow at a Fig. 6. DTA curves at a heating rate of 10 C minÀ1 for the untreated heating rate of 10 CmÀ1. The overall curve resulted of adding the muscovite in argon flow (a) and for the muscovite sonicated for 100 h DTG traces corresponding to the different fractions is plotted as a in argon flow (b) and in static air (c). 2798 L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 compound present or formed during the sonication release from the tip cup of the ultrasound probe. This Ti treatment. The chemical analysis as assayed by X-ray is present as discrete particles mixed with the mica par- fluorescence of the muscovite and biotite samples soni- ticles. The Ti particles are oxidized to rutile when heated cated for different times show similar composition, under static air remaining as independent particles except for the percentages of TiO2 that suffer a remark- mixed with the mica. While the importance of con- able increase in relation with the untreated samples. tamination during grinding of minerals has been 25,27 Thus, the TiO2 percentages change from 0.05% for previously recognized in the literature, no data are untreated sample to 5.33% in muscovite sonicated for published, up to our knowledge, on the contamination 100 h. This increase in titanium is due to the release of of solids subjected to sonication treatment. this element from the tip cup of the ultrasound equip- For biotite, sonication has also a significant effect on ment that is manufactured of Ti. The powder X-ray its thermal behavior. the DTG curves recorded in argon diffraction studied of sonicated muscovite and biotite flow in the range 900–1225 C of biotite samples before samples confirms the presence of titanium element that and after sonication are shown in Fig. 8. This tempera- is oxidized to titanium oxide, rutile, during the heating ture range has been selected because the thermal dehy- in static air (figure not shown). The muscovite sample droxylation of biotite takes place in this range. The after 100 h of sonication and heated at 1200 C(Fig. 7A) DTG curve for the untreated biotite (Fig. 8a) shows a is constituted by platelets particles constituted by Si, Al peak with a maximum at 1160 C. The weight loss cor- and K (Fig. 7B) that correspond to the muscovite responding to dehydroxylation, as would be expected material and spherical particles constituted by titanium for a magnesium containing mineral, takes place at a oxide (Fig. 7C). These data confirm that sonicated considerably higher temperature than that of an alumi- samples are contaminated by particles of Ti that is num containing mineral muscovite. After 10 h sonica-

Fig. 7. (A) Scanning electron micrograph of the muscovite sample sonicated for 100 h and heated at 1200 C. (B) EDX analysis of platelets particles marked in the micrograph as 1. (C) EDX analysis of spherical particles marked in the micrograph as 2. L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 2799

Fig. 9. DTG traces of the different fractions with different particle sizes separated by the deposition-centrifugation procedure recorded for the biotite sample sonicated for 40 h under argon flow at a heating rate of 10 CmÀ1. The overall curve resulted of adding the DTG traces corresponding to the different fractions is plotted as a solid line.

than 2 mm, while the lower temperature DTG peak (with a maximum at 1050 C) is mostly due to particles smaller than 2 mm. For biotite, unlike muscovite where particle size reduction is leveled off at 40 h sonication time, the particle size reduction takes place along the entire sonication time. This continuous particle size Fig. 8. DTG curves obtained under argon flow and a heating rate of reduction has an effect also in the thermal behavior. À1 10 C min for the biotite sample before (a) and after sonication for Thus, the lower temperature peak increases while the 10 h (b), 20 h (c), 40 h (d), 50 h (e) and 100 h (f). high temperature effect decreases with sonication time, indicating that the amount of particles with sizes smaller tion time, the DTG trace (Fig. 8b) shows that the peak than 2 mm is increasing while the amount of the larger centered at 1160 C is broader and it is also accom- ones is decreasing. panied by a shoulder at lower temperature. This The DTA and TG performed in air for the sonicated shoulder is transformed into a peak at 1052 C after 20 biotite samples (figures not shown) also indicated con- h treatment (Fig. 8c). The intensity of this second peak tamination of Ti that oxidizes during heating. This Ti increases with sonication time, while the intensity of the contamination was also detected by X-ray diffraction, high temperature peak decreases. Thus, after 100 h chemical analysis and SEM in a similar way as in the treatment time the high temperature peak has almost muscovite samples. disappeared and it is only present as a small shoulder (Fig. 8f). In order to study the contribution of the different 4. Conclusions particle sizes on the thermal behavior of the biotite sample, the same particle size separation procedure used Sonication does not produce significant structural for muscovite was applied to biotite. Fig. 9 includes for transformation in the mica samples. Additionally, for the biotite sample sonicated for 40 h, the DTG trace of the muscovite sample, it has been observed by NMR the different fractions with the overall curve resulted of that the coordination of Al and Si does not suffer adding up the contribution of all the fractions. In this modification due to the treatment. Nevertheless, the case, as it happened for muscovite, the curve resulted of particle size of the samples is drastically reduced. This adding up all the individual DTG traces present the particle size reduction produces important changes in same shape as that included in Fig. 8d for the sample the thermal behavior of these micas. For the muscovite sonicated for 40 h. Fig. 9 shows that particles larger sample, the weight losses on the DTG curve due to than 2 mm dehydroxylate at higher temperatures than dehydroxylation and the corresponding DTA endother- those smaller than 2 mm. Thus, the DTG peak at higher mic effect are broadened and shifted to lower tempera- temperature (with a maximum at 1160 C) is due to the ture, until a sonication time limit is reached at 40 h, dehydroxylation of particles with an average size larger treatments longer than 40 h do not produce any impor- 2800 L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 tant change in the TG or DTA traces. This behavior 8. Grim, R. E., Clay Mineralogy. McGraw-Hill, New York, 1968. could be correlated with the decrease of particle size up 9. Pascual, J., Zapatero, J., de Haro, M. C. J., Varona, I., Justo, A., to 40 h sonication time followed by a leveling off. The Perez-Rodriguez, J. L. and Sanchez-Soto, P. J., Porous mullite and mullite-based composites by chemical processing of kaolinite broadening of the main weight loss is related to the and aluminium metal wastes. J. Mater. Chem., 2000, 10, 1409– effect of the particle size distribution. The weight loss in 1414. the range from 300 to 450 C could be attributed to 10. Miller, J. G. and Oulton, T. D., Protrotropy in kaolinite during loosely OH in the borders of the particles generated by percussive grinding. Clays Clay Miner., 1970, 18, 313–323. the sonication treatment. 11. Cicel, B. and Kranz, G., Mechanism of montmorillonite structure degradation by percussive grinding. Clay Miner., 1981, 16, 151– For the biotite, the particle size decreases in the entire 162. sonication time range studied here, and this continuous 12. Juhasz, Z. and Somogoy, A., Grinding test with illite. Keram., change has an important effect on the biotite thermal 1984, 36, 659–662. behavior. Thus, the only weight loss in the range from 13. Pe´ rez-Rodrı´ guez, J. L., Madrid, L. and Sa´ nchez Soto, P. J., 900 to 1200 C due to dehydroxylation observed for the Effects of dry grinding on pyrophyllite. Clay Miner., 1988, 23, 399–410. untreated biotite is converted into two different weight 14. Sanchez-Soto, P. J., Wiewiora, A., Aviles, M. A., Justo, A., losses for the sonicated samples. The DTG trace shows Perez-Maqueda, L. A., Perez-Rodriguez, J. L. and Bylina, P., that one of the peak increases its intensity while the Talc from Puebla de Lillo, Spain. II. Effect of dry grinding other decreases with the sonication time. The study of on particle size and shape. Appl. Clay Sci., 1997, 12, 297– the influence of particle size distribution on the thermal 312. 15. Sanchez-Soto, P. J., Ruiz-Conde, A., Aviles, M. A., Justo, A. and behavior has shown that the low temperature peak is Pe´ rez-Rodrı´ guez, J. L., Mechanochemical effects on vermiculite due to particles smaller than 2 mm while the one at and its influence on the synthesis of nitrogen ceramics. In Cera- higher temperatures is due to particles larger than 2 mm. mic Chartingthe Future , ed. P. Vicenzeni. Techno Srl, Faenza The contamination of sonicated samples has been (Italy), 1995, pp. 1383–1390. reported, up to our knowledge, for the first time in this 16. Sanchez-Soto, P. J., Del Carmen Jimenez de Haro, M., Perez- Maqueda, L. A., Varona, I. and Perez-Rodriguez, J. L., Effects of paper. Thus, it has been observed that Ti is released dry grinding on the structural changes of kaolinite powders.J. from the sonication probe and appears as individual Am. Ceram. Soc., 2000, 83, 1649–1657. particles mixed with the sonicated particles. 17. Perez-Maqueda, L. A., Caneo, O. B., Poyato, J. and Perez- Rodriguez, J. L., Preparation and characterization of micron and submicron-sized vermiculite. Phys. Chem. Miner., 2001, 28, 61– 66. Acknowledgements 18. Suslick, K. S., Application of ultrasound to materials chemistry. MRS Bull., 1996, 74, 29–34. This publication was prepared on the basis of the 19. Perez-Rodriguez, J. L., Carrera, F., Poyato, J. and Perez- results obtained in the frame of the project MAT 2002- Maqueda, L. A., Sonication as a tool for preparing nanometric 03774 of the Ministry of Science and Technology of vermiculite particles. Nanotechnology, 2002, 13, 382–387. 20. Wiewiora, A., Perez-Rodriguez, J. L., Perez-Maqueda, L. A. and Spain. Drapala, J., Particle size distribution in sonicated high- and low- charge vermiculites. Appl. Clay Sci. (in press). 21. Kupar, P. C., Kinetics of solid-state reactions of particulate References ensembles with size distributions. J. Am. Ceram. Soc., 1973, 56, 79–81. 1. Hedrick, J. B., Mica. Am. Ceram. Soc. Bull., 1999, 78, 136–138. 22. Koga, N. and Criado, J. M., Kinetic analyses of solid-state reac- 2. Muller, F., Drits, V., Plancon, A. and Besson, G., Dehydroxyla- tions with a particle-size distribution. J. Am. Ceram. Soc., 1998, tion of Fe3+, Mg-rich dioctahedral micas: (I) structural transfor- 81, 2901–2909. mation. Clay Miner., 2000, 35, 491–504. 23. Lahiri, A. K., The effect of particle size distribution on TG. 3. Mackenzie, K. J. D., Brown, I. W. M., Cardile, C. M. and Thermochimica Acta, 1980, 40, 289–295.

Meinhold, R. H., The thermal reactions of muscovite studied by 24. Cooper, E. A. and Mason, T. O., Mechanism of La2CuO4 solid high-resolution solid-state 29Si and 27Al NMR. J. Mater. Sci., state powder reaction by quantitave XRD and impedance spec- 1987, 22, 2645–2654. troscopy. J. Am. Ceram. Soc., 1995, 78, 857–864. 4. Guggenheim, S., Chang, Y.-H. and Koster van Groos, A. F., 25. Pe´ rez-Maqueda, L. A., Pe´ rez-Rodrı´ guez, J. L., Scheiffele, G. W., Muscovite dehydroxilation: high-temperature studies. Am. Justo, A. and Sa´ nchez-Soto, P. J., Thermal analysis of ground Mineralogist, 1987, 72, 537–550. kaolinite and pyrophyllite. J. Therm. Anal., 1993, 39, 1055– 5. Ovadyahu, D., Yariv, S., Lapides, I. and Deutsch, Y., Mechan- 1067. ochemical adsorption of phenol by tot swelling clay minerals II 26. Pe´ rez-Rodrı´ guez, J. L. and Sa´ nchez-Soto, P. J., The influence of simultaneous DTA and TG study. J. Therm. Anal. Calorim., dry grinding on the thermal behaviour of pyrophyllite. J. Therm. 1998, 51, 431–447. Anal., 1991, 37, 1401–1413. 6. Matteazi, P. and Le Caer, G., Room temperature mechan- 27. Pe´ rez-Rodrı´ guez, J., Pe´ rez-Maqueda, L., Justo, A. and Sa´ nchez- osynthesis of carbides by grinding of elemental powders. J. Am. Soto, P., Influence of grinding contamination on high tempera- Ceram. Soc., 1991, 74, 1382–1390. ture phases of kaolinite. Ind. Ceram., 1992, 12, 109–113. 7. Alcala, M. D., Gotor, F. J., Perez-Maqueda, L. A., Real, C., Dianez, 28. Wiewiora, A., Sanchez-Soto, P. J., Aviles, M. A., Justo, A., M. J. and Criado, J. M., Constant rate thermal analysis (CRTA) as Perez-Maqueda, L. A., Perez-Rodriguez, J. L. and Bylina, P., a tool for the synthesis of materials with controlled texture and Talc from Puebla de Lillo, Spain. I. XRD study. Appl. Clay Sci., structure. J. Therm. Anal. Calorim., 1999, 56, 1447–1452. 1997, 12, 233–245. L.A. Pe´rez-Maqueda et al. / Journal of the European Ceramic Society 24 (2004) 2793–2801 2801

29. Pe´ rez-Maqueda, L. A., Franco, F., Avile´ s, M. A., Poyato, J. and 32. Sanz, J. and Serratosa, J. M., Si-29 and Al-27 high-resolution Pe´ rez-Rodrı´ guez, J. L., Eﬀect of sonication on particle-size dis- MAS–NMR spectra of phyllosilicates. J. Am. Chem. Soc., 1984, tribution in natural muscovite and biotite. Clays Clay Miner., 106, 4790–4793. 2003, 51, 701–708. 33. Wiewio´ ra, A. and Weiss, Z., X-ray powder transmission dif- 30. Jackson, M. L., Soil Chemical Analysis—Advanced Course, 2nd fractometry determination of mica polytypes: method and appli- edn. Published by the author, Madison, WI, 1975. cation to natural samples. Clay Miner., 1985, 20, 231–248. 31. Sanz, J. and Serratosa, J. M., Distinction of tetrahedrally and 34. Mackenzie, R. C., Simple phyllosilicates based on gibbsite and octahedrally coordinated Al in phyllosilicates by NMR-spectro- brucite-like sheets. In Diﬀerential Thermal Analysis, Vol. 1, ed. scopy. Clay Minerals, 1984, 19, 113–115. R. Mackenzie. Academic Press, London, 1970, pp. 498–537.